高阶统计分析的m序列检测新方法

A new method for m-sequence detection using higher-order statistics

为了提高直扩信号中m序列的检测效率,研究高阶统计分析(HOS)理论的m序列检测问题.在理论上,重新对m序列三阶相关函数的形式作了正向和反向的区分.由m序列三阶相关函数峰值求得本原多项式的一般推导方法,并结合多项式求最大公因式的矩阵斜消变换理论,提出利用高阶统计分析理论求m序列本原多项式的规律算法,仿真结果验证了此方法有效可行.同时,分析了高斯噪声信道下此算法的性能,仿真结果表明,此算法在-10 dB时的检测概率接近70%.

In order to overcome the m-sequence detection problem in direct sequence （DS） signals, research was carried out on higher-order statistical （HOS） signal processing. The triple correlation function （TCF） was theoretically redefined with forward and backward forms. The idea of deriving a primitive polynomial from a particular TCF＇s peak coordinates was combined with the matrix＇s oblique elementary theory to evaluate the greatest common factor of a given polynomial. Use of the regular algorithm for deriving primitive polynomial was then proposed. Resuits from a simulation proved the algorithm feasible and effective. At the same time, performance of the algorithm with additive white Gaussian noise （AWGN） was analyzed. Simulation data indicated the detection efficiency was close to 70% at - 10dB.